Ok, I just have realized that the Samsug NVMe driverpack supports 32bit and 64bit Operating Systems.
I suspect, that the additional Samsung SCSI Filter Driver named nvmeF.sys is not usable with Windows XP (32/64bit). It may be the same thing as with the Intel SCSI Filter Driver named iaStorF.sys, which cannot be handled by XP.
Fernando,
seems a new v1.1 pack Samsung NVMe driver released.
A difference with 1.0?
I am always with Microsoft drivers, because with Samsung
The temperature of the SSD is higher (I do not know what value is really good …)
The order of the drive are reversed (the C drive is not disk 0 but disk 1)
Nothing important but with the Microsoft drivers, I do not have all these "defects" and that is almost the same for performances
A test as if with these new drivers, there is no "bug" if we install them from the .inf (Device Manager)
What means "seems"?
Where did you read/hear about the release? Is there any download link? If yes, where?
I have installed and used v1.4.7.17.
All is working fine. I ignore what are the change.
Here is a link:
http://www.touslesdrivers.com/index.php?v_page=3&v_code=6050
@ all:
As soon as I am able to extract the "pure" NVMe driver files, I will present them within the start post of this thread.
C drive remains in "Disk 0" or as with older drivers, it is disk 1 ?
Thanks
I ignore what is the Disk 0 problem, i use this SSD as a non system storage media.
Here are some screens:
I ignore why Random Read Score is only 137174 instead of about 300 000 expected…
It’s not really a problem, Windows 10 is on the 950 Pro (drive C) for me
With the MS driver, drive C is the disk 0, with the Samsung driver is the disk 1
For example in the Task Manager, Performance tab, we can see it
Nothing serious, but I prefer to have the C drive => Disk 0
I understand now your question.
On my PC you can see on the screencopy my System disk is "Disque 3".
This is not problem (for me)
I observed that W10 seems to attribute a different Disk number each time you reconfigure something in the PC system (i.e. add/remove a new SSD for doing a test, or for driver controller modification as you report, etc…), without memorizing what is the original System Disk number. I ignore why W10 does that. If it is working fine it is sufficient for me.
Hi Fernando,
Samsung released new NVMe driver 1.4.7.17 WHQL, please see attachment. 
Regards
morpheus67
Samsung_nvme_1.4.7.17.zip (1.47 MB)
Update of the start post
Changelog:
Hello
I test the latest Samsung drivers, I installed the .exe and I have not nvmeF.sys file (I see they are available in your archives "Drivers")
This file is present for you?
I am running Windows 10
Thanks
EDIT: So I tried installing the Samsung drivers by the device manager (.inf), and I do not have a USB detection problems (not the nvmeF.sys file in the folder where is located the files drivers)
I will other tested on a fresh install (and integrating drivers) to be sure not to have problems
@ tistou77:
It may be the same thing as with the Intel RST(e) drivers: The Windows Operating Systems from Win8 up do not need the SCSI Filter Driver named nvmeF.sys resp. iaStorF.sys, because they natively do support the UNMAP command, which is responsable for letting the Trim command passing through the related Storage Controller into the SSD.
Ok, this is maybe this nvmeF.sys file that cause problems with USB key
Update of the start post
Changelog:
Update of the start post
Changelog:
Hi. for me, NO trim is not supported.
My System:
Windows 10 Pro
Marvell 9230 Controller with 2 mSATA ports and 2 SATA3 ports
Firmware Version: 2.3.1063 (I updated the firmware today, hoping it will bring trim support with the new drivers)
trimcheck-0.7.exe used to verify trim support.
Sad but true, back to MSAHCI
the “old” (and maybe faster) >64bit AMD AHCI Driver v1.2.1.359 WHQL for Win7 x64< does not boot my windows 7. it’s stuck at windows starting screen. the “new” one does booting completely. so i have to choose between standard windows ahci driver or the “new” one. i did a benchmark with my sshd. yeah, i know, it’s difficult to benchmark a sshd because of judgement of often used files layed on the flash area of the sshd. but the results are quite equal. some values were faster, the other slower…all tests are running on SATA II
AMD “new”:
standard windows ahci driver:
EDIT: here are some pics of another benchmark
standard:
amd “new”:
